From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from vsmx002.dclux.xion.oxcs.net (vsmx002.dclux.xion.oxcs.net [185.74.65.108]) (using TLSv1.2 with cipher ADH-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by lists.bufferbloat.net (Postfix) with ESMTPS id DD0E43B2A4 for ; Wed, 31 Aug 2022 17:09:20 -0400 (EDT) Received: from proxy-1.proxy.oxio.ns.xion.oxcs.net (proxy-1.proxy.oxio.ns.xion.oxcs.net [83.61.18.4]) by mx-out.dclux.xion.oxcs.net (Postfix) with ESMTPA id 8309E8C58D1 for ; Wed, 31 Aug 2022 21:09:18 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=simple/simple; d=dclux.xion.oxcs.net; s=mail1; t=1661980159; bh=4ChxOHxTFolD6/O+DLABDONpRoIbsrqEZJ9JATD78Q4=; h=Date:From:To:In-Reply-To:References:Subject:From; b=Yo6O+ieYIi5xRdTmHVNG1NMSA5r/Yaki36THzmvWw+CI6sr7lhcZkpwiuyRIABJMH wO33i2qoBGfRaxuCXYWgFg0AYMGuaqqiA7kVwPVCruLuIn2tS9UVTp7VuKx1okgRSs q/urB2H9XLOI9tkFfxyQHZ7t+kt9NmiKVCdri3a8Wd2YmPwlUtLWTha8jLEXR/5BbC wY3izB/GRH4/Ya/BSxTy0SdXL7zse/UHsrxaHY6DIUZ+B8kv+GzDCBGHnCiRVXzyol AUs4a/rCrJbTjC2Dbpc/niEoQyFBQM9C/y//b7tsMqfxb3j8jtfCJX2grQlhizVoUO YlBKGCej4ViEQ== Date: Wed, 31 Aug 2022 23:09:01 +0200 From: Mike Puchol To: "David P. Reed via Starlink" Message-ID: In-Reply-To: <1661977461.578310669@apps.rackspace.com> References: <1661977461.578310669@apps.rackspace.com> X-Readdle-Message-ID: abd72ac2-d5cb-4570-91d6-a558e4e830b0@Spark MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="630fcdfd_257130a3_3328" X-VadeSecure-Status: LEGIT X-VADE-STATUS: LEGIT Subject: Re: [Starlink] CDNs in space! X-BeenThere: starlink@lists.bufferbloat.net X-Mailman-Version: 2.1.20 Precedence: list List-Id: "Starlink has bufferbloat. Bad." List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 31 Aug 2022 21:09:21 -0000 --630fcdfd_257130a3_3328 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable Content-Disposition: inline I will speak from the side of a tiny WISP in Kenya (15k customers) who wa= s offered a free OCA by Netflix, after they noticed certain levels of dem= and from our AS. This OCA is a 1U server with 144TB of storage. It now supplies 10% of our= customer traffic. Not having to pull all that traffic from eg South Africa is good for us, = good for Netflix, and good for the customer. I have had the firm belief, for over a year now,=C2=A0=C2=A0that SpaceX w= ill field OCA-type content caches in the satellites. They add more satell= ites than gateway capacity, so having a significant portion of traffic de= livered over ISL from CDNs in some satellites (you don=E2=80=99t need the= m all to have a cache) is a clear benefit. The critical problem to solve is cache hit ratio, considering content wan= ted from multiple markets and regions. My other prediction is that eventually they will have to move to optical = gateway, as the terahertz gap will get in the way. Best, Mike On Aug 31, 2022, 22:24 +0200, David P. Reed via Starlink , wrote: > Having looked into this a lot, CDNs don't account for very much Interne= t traffic. There's a lot of marketing propaganda out there that suggests = this might be true, but when you try to track down the source, it's almos= t always from an old slide deck presented by a company that sells CDN ser= vers. > That doesn't mean that CDN servers don't help a fair bit, but no, today= 's Internet wouldn't fall apart if we didn't have CDN servers withint 2 m= sec. of the edge. > Also, CDN's need to be BIG to hold all the videos that people might cho= ose to watch at any particular time. > So I'm just pointing out that the business case for CDN's in space to m= erely solve Starlink's potential issues is probably not great. Maybe 10 y= ears from now. Maybe never. The idea that everyone watches TV and the sam= e few seconds of content of a few shows that are extraordinarily popular = - well, that dog don't hunt. It doesn't justify multicast either. > So let's improve the discussion here. The Internet, for the forseeable = future, at the edge, is unicast. > CDN's reduce costs for big companies like Netflix, not because they are= =22close to the watcher=22 but because it is easier to not have everythi= ng centralized in a single point of failure and needing a huge pipe becau= se all customers aggregate demand into their central HQ. > There are some advantages at the top level peering in distributing CDN = sites into the networks of various access providers (like Comcast) becaus= e you can get cheaper pricing. > But lets think about CDNs in space. They don't deal with the actual bot= tleneck in each satellite to ground. The uplink from central ground stati= ons has LOTS of capacity, it won't be the bottleneck if Starlink or any o= ther satellite service for that matter balances its design. > And if they get inter-sat links (laser or maybe R=46, if the interferen= ce problem with other satellites is resolved at the WRC or other internat= ional body that regulates R=46 in space) working, where are the CDNs goin= g to maintain state=3F The reachable ones for any satellite in the conste= llation will be moving relative to the user's dishy, unless it the CDN is= in geosync orbit). > Starlink isn't a media company. It doesn't want to own all the content,= or even host all the content. > So I'm pretty darn skeptical=21 > At the moment, Starlink is a =22last mile=22 service in a virtual sense= . It connects from the public internet at some pretty high performance po= int of presence, hopefully has a pretty good latency, and doesn't let its= users saturate any satellite (lest huge queueing delay build up). Given = that, anywhere there is a ground station the public fiber backbone can re= ach all the CDNs. > It's hard to capitalize a last mile fiber service. Each home passed cos= ts about =242000, if we want to connect up (that includes almost all actu= al rural residences, but not all the uninhabited parts of the planet). =46= or whatever reason, the telcos want no one else to put in fiber, but also= don't see much profit in extending their coverage, either, because they = aren't allowed to charge for all the content like they used to. =46iber's= great advantage is that it has extremely low operational expense. > This is what Starlink fills in for. It has lower capex, but HUGE opex. = It can't perform or scale very well compared to the asymptotic fiber solu= tion. So it has a pretty good short-term competitive position. And they'v= e gotten very smart in finding early customers who are eager to buy in. > However, it's important to realize that what drives this all is how the= Internet is used and who needs it where, and how much they are willing t= o pay. > One thing is clear - Starlink isn't the Internet of the future. It's fi= lling a niche (a large one, but a niche). > The mistake Motorola made with Iridium was in not realizing that cellul= ar telephony was in all respects a better answer, and would be cheaper, t= oo. People did talk about putting CDNs in space with Iridium, too. > But one needs to understand what CDNs are useful for, and really unders= tand what part of the problem that is. > =5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F= =5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F > Starlink mailing list > Starlink=40lists.bufferbloat.net > https://lists.bufferbloat.net/listinfo/starlink --630fcdfd_257130a3_3328 Content-Type: text/html; charset="utf-8" Content-Transfer-Encoding: quoted-printable Content-Disposition: inline
I will speak from the side of a tiny WISP in Kenya = (15k customers) who was offered a free OCA by Netflix, after they noticed= certain levels of demand from our AS.&=23160;

This OCA is a 1U server with 144TB of storage. It now supplies 10% of our= customer traffic.

Not having to pull all that traffic from eg South Africa is good for us, = good for Netflix, and good for the customer.

I have had the firm belief, for over a year now,&=23160;&=23160;that Spac= eX will field OCA-type content caches in the satellites. They add more sa= tellites than gateway capacity, so having a significant portion of traffi= c delivered over ISL from CDNs in some satellites (you don=E2=80=99t need= them all to have a cache) is a clear benefit.&=23160;

The critical problem to solve is cache hit ratio, considering content wan= ted from multiple markets and regions.

My other prediction is that eventually they will have to move to optical = gateway, as the terahertz gap will get in the way.&=23160;

Best,

Mike
On Aug 31, 2022, 22:24 +0200, David= P. Reed via Starlink <starlink=40lists.bufferbloat.net>, wrote:

Having= looked into this a lot, CDNs don't account for very much Internet traffi= c. There's a lot of marketing propaganda out there that suggests this mig= ht be true, but when you try to track down the source, it's almost always= from an old slide deck presented by a company that sells CDN servers.

That d= oesn't mean that CDN servers don't help a fair bit, but no, today's Inter= net wouldn't fall apart if we didn't have CDN servers withint 2 msec. of = the edge.

Also, = CDN's need to be BIG to hold all the videos that people might choose to w= atch at any particular time.

So I'm= just pointing out that the business case for CDN's in space to merely so= lve Starlink's potential issues is probably not great. Maybe 10 years fro= m now. Maybe never. The idea that everyone watches TV and the same few se= conds of content of a few shows that are extraordinarily popular - well, = that dog don't hunt. It doesn't justify multicast either.

So let= 's improve the discussion here. The Internet, for the forseeable future, = at the edge, is unicast.

CDN's = reduce costs for big companies like Netflix, not because they are =22clos= e to the watcher=22 but because it is easier to not have everything centr= alized in a single point of failure and needing a huge pipe because all c= ustomers aggregate demand into their central HQ.

There = are some advantages at the top level peering in distributing CDN sites in= to the networks of various access providers (like Comcast) because you ca= n get cheaper pricing.

But le= ts think about CDNs in space. They don't deal with the actual bottleneck = in each satellite to ground. The uplink from central ground stations has = LOTS of capacity, it won't be the bottleneck if Starlink or any other sat= ellite service for that matter balances its design.

And if= they get inter-sat links (laser or maybe R=46, if the interference probl= em with other satellites is resolved at the WRC or other international bo= dy that regulates R=46 in space) working, where are the CDNs going to mai= ntain state=3F The reachable ones for any satellite in the constellation = will be moving relative to the user's dishy, unless it the CDN is in geos= ync orbit).

Starli= nk isn't a media company. It doesn't want to own all the content, or even= host all the content.

So I'm= pretty darn skeptical=21

At the= moment, Starlink is a =22last mile=22 service in a virtual sense. It con= nects from the public internet at some pretty high performance point of p= resence, hopefully has a pretty good latency, and doesn't let its users s= aturate any satellite (lest huge queueing delay build up). Given that, an= ywhere there is a ground station the public fiber backbone can reach all = the CDNs.

It's h= ard to capitalize a last mile fiber service. Each home passed costs about= =242000, if we want to connect up (that includes almost all actual rural= residences, but not all the uninhabited parts of the planet). =46or what= ever reason, the telcos want no one else to put in fiber, but also don't = see much profit in extending their coverage, either, because they aren't = allowed to charge for all the content like they used to. =46iber's great = advantage is that it has extremely low operational expense.

This i= s what Starlink fills in for. It has lower capex, but HUGE opex. It can't= perform or scale very well compared to the asymptotic fiber solution. So= it has a pretty good short-term competitive position. And they've gotten= very smart in finding early customers who are eager to buy in.

Howeve= r, it's important to realize that what drives this all is how the Interne= t is used and who needs it where, and how much they are willing to pay.

One th= ing is clear - Starlink isn't the Internet of the future. It's filling a = niche (a large one, but a niche).

The mi= stake Motorola made with Iridium was in not realizing that cellular telep= hony was in all respects a better answer, and would be cheaper, too. Peop= le did talk about putting CDNs in space with Iridium, too.

But on= e needs to understand what CDNs are useful for, and really understand wha= t part of the problem that is.

=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F= =5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F
Starlink mailing list
Starlink=40lists.bufferbloat.net
https://lists.bufferbloat.net/listinfo/starlink
--630fcdfd_257130a3_3328--